Removal of the SWPC scaling factor for GOES data prior to GOES-16

[Vancanneyt Sander]

It was brought to our attention that there is a notable change between operational and scientific data from the GOES satellites. The operational data prior to GOES-16 had a scaling factor applied to it by the Space Weather Prediction Center (SWPC) to match those from GOES-7. But by doing that, the real irradiances are different and don't match the true physical units of W/m2. All science quality data going from GOES-8 to GOES 15 has been modified to remove the SWPC scaling. Because of the scaling factors in operational data, flare indices for the operational data is 42% smaller than for the science quality data (for example an X2.5 class flare prior to GOES-16 is an X3.6 class solar flare.

Some history:

At SpaceWeatherLive we like to be correct whenever we talk about solar flares, so we have made some big changes. All operational data that is in our archive prior to the day that GOES-16 became the primary satellite is now modified to remove the SWPC scaling factor. The solar flare database has been updated with +500 manual corrections of solar flares to rule out manmade errors from the past so that the all solar flares prior to GOES-16 (38.000 solar flares) could be updated with their true physical units based on the science quality data. 8.373 dates in the archive have received edits to the number of C, M, X-class events on each day and the modified background flux with their true physical units. In reports and news items prior to GOES-16, an informational banner will be shown so you are informed that the displayed solar flares in the news item or report have a scaling factor applied to them. 

A special thanks to @Jesterface23 for providing dumps of the science quality data! This has been a true help in doing this major update.

PS.: due to caching at our Content Delivery Network, it's possible that for some dates you don't see the changed data yet, same for the solar flare top lists and solar cycle progression page. This an take up to 30 days as it is archived content with long caching period.

[Marcel de Bont]

This is a huge task and an incredible amount of work has been done by @Vancanneyt Sander to make these changes accurate and a reality. A huge shout out and special thanks to @Jesterface23 as well for his contributions! Another step towards making the SpaceWeatherLive archive as complete as possible.

[Sam Warfel]

Cool, I only learned about that recently thanks to the knowledgeable people here responding to my recent thread about the discrepancy. 
I’m glad to hear it’s been updated!

[hamateur 1953]

Way cool. My favorite region 10486 is now officially an X-40 or a Y-4 if @Newbieweighs in too. Haha.  

[libmar96]

Is this applied also to alerts about incoming radio blackout? We just got a R1 alert with a note about current value of M3.33, however the graph never got that exact value - after M2.54 there was a jump to M3.38. It's actually a close value, but not exact and made me think if the correction was actually applied there as well.

[Vancanneyt Sander]

16 minuten geleden, libmar96 zei:

Is this applied also to alerts about incoming radio blackout? We just got a R1 alert with a note about current value of M3.33, however the graph never got that exact value - after M2.54 there was a jump to M3.38. It's actually a close value, but not exact and made me think if the correction was actually applied there as well.

Current GOES satellites don’t have the correction factor. It only applies to the satellites prior to GOES-16. Currently GOES-18 is primary satellite.

keep in mind that our alerts get send by our server when data is pulled from the satellite feed, what you see on the site/app can be few minutes behind due to caching of data to be able to handle the loads of visitors that come after an alert. If we wouldn’t cache, we would go down after each alert which isn’t what we want 😉

[Philalethes]

On 12/31/2023 at 6:51 PM, hamateur 1953 said:

Way cool. My favorite region 10486 is now officially an X-40 or a Y-4 if @Newbieweighs in too. Haha.  

Imagine if we ever got to see a Z-flare, heh. With the new GOES scaling, the Carrington flare is adjusted from the estimate of X45 to X64 in current terms, as mentioned here:

Quote

For example, the peak SXR classes for the 1 September 1859 (X45 ± 5) and 774 AD (~ X285 ± 140) events will increase to X64.4 ± 7.2 and ~ X410 ± 200, respectively, while that for the largest possible solar flare changes from X180 (− 100, + 300) to X260 (− 140, + 430).

And those other estimates are also quite wild, although naturally far more speculative; but I'd certainly love to see a Z-flare myself.